Violence Against or Obstruction of Health Care in Somalia in 2020 - Insecurity Insight
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Insecurity
Insight
Data on People in Danger
Violence Against
or Obstruction of
Health Care in
Somalia in 2020Letter from the Chair
The fifth anniversary of the United Nations (UN) Security Council’s Resolution
2286 on the protection of health care comes at a time of unceasing violence
inflicted on hospitals, clinics, ambulances and health workers. As this report
shows, the number of health workers reported killed in conflict settings rose to
185 in 2020, up from 167 and 150 in 2018 and 2019, respectively. It was a rare
conflict where escalation in fighting was not associated with a corresponding
upsurge in violence against health care of some kind.
During the five years since the UN resolution was adopted, 14 conflicts have seen more than 50
reported incidents of violence against health care, eight conflicts have seen more than 100 such
incidents, five more than 200, and four more than 300 incidents apiece. This is probably an
undercount, and the real numbers are likely to be much higher. Violence against health care is
continuing in 2021.
The reasons for the violence are variable and sometimes complex, but the explanation for
continuing impunity is not: states have failed to fulfil their commitments to take action – individually
or as part of an international effort – to prevent violence against health care or hold the perpetrators
accountable. Consider these questions regarding implementation actions found in the resolution
itself or the UN Secretary-General’s recommendations for implementation:
Did member states ensure that their militaries ‘integrate practical measures for the
protection of the wounded and sick and medical services into the planning and
conduct of their operations’? – No.
Did member states adopt domestic legal frameworks to ensure respect for health
care, particularly excluding the act of providing impartial health care from punishment
under national counter-terrorism laws? – No.
Did member states engage in the collection of data on the obstruction of, threats
against and physical attacks on health care? – No.
Did member states undertake ‘prompt, impartial and effective investigations within
their jurisdictions of violations of international humanitarian law’ in connection with
health care and, ‘where appropriate, take action against those responsible in
accordance with domestic and international law?’ – No.
Did the Security Council refer cases where there is evidence of war crimes in
connection with violence against health care in Syria and elsewhere to the
International Criminal Court? – No.
Were all states found by the Special Representative of the Secretary-General on
Children in Armed Conflict to have engaged in violence against hospitals listed in
the annex to the Secretary-General’s annual report on children in armed conflict?
– No.
Did member states that sell arms that have been used to inflict violence on health
care cease those sales? – No.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 1Letter from the Chair Non-state armed groups, many of which profess their commitment to abide by international law, have also abdicated their responsibilities. Only three have signed the Geneva Call’s Deed of Commitment to Health Care. This compares to more than 50 non-state armed groups that have agreed to forgo the use of antipersonnel landmines and 25 that have agreed not to use child soldiers. Why the inaction? Militaries do not change their operational procedures if there are few demands on them to do so. Laws are not reformed when counter-terrorism priorities pay little regard to international law. Arms sales are huge moneymakers and a valued way of achieving policy goals without direct military involvement. Investigations and accountability are inconvenient in a conflict. At the UN, the very structure of the Security Council – especially the veto power of its five permanent members – has become an excuse for failure. If governments are to do what they have committed to – i.e. protect health workers, health facilities, and transport from being targeted and attacked – both pressure and accountability are urgently needed. To that end, the UN Secretary-General has the power to and should report every year on what each UN member state has done and not done to carry out the purposes of Resolution 2286. This form of accountability can also be advanced by the appointment of a special rapporteur or special representative to submit reports thematically and on countries to assess their response to the requirements of Resolution 2286. Most of all, the public health, nursing, and medical communities must demand that political leaders move beyond declarations, meetings, and pallid measures and take concrete steps to ensure that health workers and the sick and wounded who need care are properly protected. It is long overdue for the important commitments of UN Resolution 2286 to be more than hollow words. All those who care about protecting health care in situations of conflict must take meaningful and concrete steps to make real these essential promises to those who risk their lives to safeguard the health and well-being of populations in their care. Len Rubenstein Chair, Safeguarding Health in Conflict Coalition NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 2
SOMALIA
REPORTED INCIDENTS AND MOST COMMONLY REPORTED CONCERNS
10 11 6
REPORTED INCIDENTS HEALTH WORKERS KIDNAPPED HEALTH WORKERS KILLED
Source: 2020 SHCC Health Care Somalia Data
OVERVIEW
The SHCC identified ten incidents of violence against or obstruction of health care in Somalia in
2020, compared to 12 such incidents in 2019. Six health workers were killed in these incidents and
11 kidnapped.
This factsheet is based on the dataset 2020 SHCC Health Care Somalia Data, which is available on
the Humanitarian Data Exchange (HDX).
Violence against or obstruction of health care in Somalia in 2020
Health workers were killed and kidnapped in the vicinity of health facilities and while delivering
medical supplies, or carrying out a health awareness campaign. Al-Shabaab fighters were often
named as perpetrators of these incidents.
In May, masked perpetrators kidnapped seven Somali health workers from a health facility run by
the local health NGO in the Middle Shabelle region. Their bodies were found the next day.1
1 Insecurity Insight. Safeguarding Health in Conflict Coalition 2021 Report Dataset: 2020 SHCC Health Care Somalia Data.
Incident number 557.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 3Methodology
This eighth report of the Safeguarding Health in Conflict Coalition (SHCC) covers 43 countries
and territories and provides details on incidents of threats and violence against health care in 17
countries and territories experiencing conflict in 2020. We referred to the Uppsala Conflict Data
Program (UCDP) 2 to determine if a country is considered to have experienced conflict in 2020,
and of these countries, we included those that had experienced at least one incident of violence
against or obstruction of health care in 2020. We discuss the 14 countries with more than 15
reported incidents in separate chapters, and the other three countries with less than 15 reported
incidents in paragraphs. Twenty-six other countries are included in the total counts, but are not
discussed in detail. Fourteen of the countries and territories covered in factsheets in 2020 were
included in factsheets in 2019. For the 2020 report, Azerbaijan, Mexico and Mozambique were
added, while Egypt, Ethiopia, Iraq, Pakistan, Sudan and Ukraine do not have country chapters in 2020.
The report uses an event-based approach to documenting attacks on health care, referred to as
‘incidents’ throughout the report. To prepare this report, event-based information from multiple
sources was cross-checked and consolidated into a single dataset of recorded incidents that
were coded using standard definitions. The full 2020 data cited in this report can be accessed via
Attacks on Health Care in Countries in Conflict on Insecurity Insight’s page on the Humanitarian
Data Exchange (HDX). The data for the 17 countries is made available as individual datasets. The
links are provided in the individual country profiles.
Definition of attacks on health care
The report follows the WHO’s definition of an attack on health care: ‘any act of verbal or physical
violence, threat of violence or other psychological violence, or obstruction that interferes with the
availability, access and delivery of curative and/or preventive health services’. In this report,
however, we do not use the word ‘attack’, but rather ‘incident’ or ‘incident of violence’, because
the word ‘attack’ is often interpreted to convey intent, whereas many reported incidents result
from indiscriminate or reckless behaviour/actions, but otherwise meet the WHO definition.
This report focuses on incidents of violence against health care in the context of armed conflict,
non-state conflict or one-sided violence, as defined by UCDP, while the WHO focuses on attacks
during emergencies.
In accordance with the WHO’s definition, incidents of violence against health care can
include bombings, explosions, looting, robberies, hijackings, shootings, gunfire, the
forced closure of health facilities, the violent searching of health facilities, fire, arson, the
military use of health facilities, the military takeover of health facilities, chemical attacks,
cyber attacks, the abduction of health workers, the denial or delay of health services,
assaults, forcing staff to act against their ethical principles, executions, torture, violent
demonstrations, administrative harassment, obstruction, sexual violence, psychological
violence and threats of violence.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 4Methodology These categories have been included insofar as they were reported in sources. However, some forms of violence, such as psychological violence, blockages of access or threats of violence, are rarely reported. We also record incidents of violence against patients in health facilities when references to the effects of violence on patients are included in descriptions of incidents. However, the impact of incidents of violence against patients is much broader and complex than individual incidents and cannot be accurately documented through event-based monitoring. Definition of conflict The SHCC report covers three types of conflict as defined by the UCDP:3 • State-based armed conflict is defined as ‘a contested incompatibility that concerns government and/or territory where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle-related deaths in one calendar year year’. • Non-state conflict is defined as ‘[t]he use of armed force between two organized armed groups, neither of which is the government of a state, which results in at least 25 battle-related deaths in a year’. • One-sided violence is defined as ‘[t]he deliberate use of armed force by the government of a state or by a formally organized group against civilians which results in at least 25 deaths in a year’. A country is included in the SHCC report if it is included on the UCDP list of one of the three types of conflict4 and if we identified at least one attack on health care perpetrated by a conflict actor, which for the purposes of this report is defined as a person affiliated with organized actors in conflict, which can be armed conflict, non-state conflict or one-sided violence as defined by the UCDP. Interpersonal violence and violence by patients against health care providers are not included in this report, even when they occurred in conflict-affected countries. In 2020 violence against specific public health programmes, such as polio vaccinations campaigns or the Ebola and COVID-19 responses, were only included when (a) the perpetrator was a member of a party to a conflict, and (b) available evidence suggested that the incident occurred either in the context of a contested incompatibility of territory or as one-sided act of violence by security forces included on the UCDP list of countries with more than 25 reported deaths from one-sided violence attributed to security forces. This is an important difference to the inclusion criteria used in the 2019 report, where all incidents that occurred in the conflict-affected eastern Democratic Republic of the Congo (DRC) in the context of the tenth Ebola response were included, even when there was not enough detail to determine whether the perpetrators were linked to a recognized conflict party or may have originated from local communities. Throughout 2020 the SHCC also monitored violence triggered by the COVID-19 pandemic. COVID-19-related threats and violence against health care are only included in the 2020 SHCC report when the incidents met the strict conflict-related inclusion criteria in relation to the country being included in one of the three UCDP lists, and the perpetrator and context of the incident were directly related to conflict, as outlined above. NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 5
Methodology
Inclusion of incidents
We included only the incidents that met the inclusion criteria for types of conflicts and perpetrators,
and for these we included the following types of incidents and details in the report dataset:
• incidents affecting health facilities, recording whether they were destroyed, damaged, looted
or occupied by armed individuals/groups;
• incidents affecting health workers, recording whether they were killed, kidnapped, injured,
assaulted, arrested, threatened or experienced sexual violence (when available, we recorded
the number of affected patients, although we acknowledge the likely serious underreporting
of these figures);
• incidents affecting health care transport, recording whether ambulances or other official
health care transport were destroyed, damaged, hijacked/stolen or stopped/delayed; and
• incidents recorded by the WHO Surveillance System of Attacks on Healthcare (SSA) for the ten
countries included in the system if the WHO confirmed the incidents.
Key definitions
Health worker: Refers to any person working in a professional or voluntary capacity in the
provision of health services or who provides direct support to patients, including
administrators, ambulance personnel, community health workers, dentists, doctors,
government health officials, hospital staff, medical education staff, nurses, midwives,
paramedics, physiotherapists, surgeons, vaccination workers, volunteers or any other
health personnel not named here.
Health worker affected: Refers to incidents in which at least one health worker was killed,
injured, kidnapped or arrested, or experienced sexual violence, threats or harassment.
Health facility: Refers to any facility that provides direct support to patients, including
clinics, hospitals, laboratories, makeshift hospitals, medical education facilities, mobile
clinics, pharmacies, warehouses or any other health facility not named here.
Health facility affected: Refers to incidents in which at least one health facility was
damaged, destroyed or subjected to armed entry, military occupation or looting.
Health transport: Refers to any vehicle used to transport any injured or ill person or woman
in labor to a health facility to receive medical care.
Health transport affected: Refers to incidents in which at least one ambulance or other
health transport was damaged, destroyed, hijacked or delayed with or without a person
requiring medical assistance on board.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 6Methodology
Sources
The aim of this report is to bring together known information on attacks on health care from
multiple sources. Access to sources differs among countries, and each source has its own
strengths and weaknesses. There are some differences in the definitions of what constitutes
attacks on health care used by the different sources that were used to compile the SHCC dataset.
Each source introduces unique reporting and selection biases, which are discussed below.
To identify incidents that meet the inclusion criteria, we used six distinct sources that provide a
combination of media-reported incidents and incidents reported by partners and network
organizations:
1. information included in Insecurity Insight’s Attacks on Health Care Monthly News
Briefs,5 which provide a combination of media sources and publicly shared information
from partner networks, such as the Aid Worker Security Database (AWSD) 6 for global
data from international aid agencies coordinating health programmes; Airwars7 and
the Syrian Network for Human Rights (SNHR) 8 for data on Syria; the Civilian Impact
Monitoring Project (CIMP) 9 for data on Yemen; and databases such as that of the
Armed Conflict Location & Event Data Project (ACLED);10
2. information provided by Medical Aid for Palestinians (MAP) 11 for incidents in the
occupied Palestinian territories (oPt);
3. information provided by SHCC member Syrian American Medical Society (SAMS)
Foundation12 for incidents in Syria;
4. information from the WHO SSA on 11 countries: Afghanistan, Burkina Faso, the DRC,
Libya, Mali, Myanmar, Nigeria, the oPt, Somalia, South Sudan and Yemen (information
from the SSA represents approximately one-third of the data gathered for this report);
and
5. information on Afghanistan from 74 WHO SSA reported incidents (but we were not
able to compare the individual reports to meaningfully combine the data).
Coding principles
The general theory and principles of event-based coding were followed, and care was taken not
to enter the same incident more than once. The standard coding principles are set out in the
SHCC Overview Data Codebook. Please see www.insecurityinsight.org/projects/healthcare/shcc
for full details of SHCC coding and annexes.
Coding the perpetrator and context of attacks on health care can inform the development of
preventive strategies and mitigation measures that reduce the incidence and impact of attacks
and support accountability processes. Because it is rarely possible to know a perpetrator’s
motive(s), we relied on the context identified in the incident descriptions and coded the
intentionality of the attacks from these descriptions to the extent possible.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 7Methodology Inclusion and coding of SSA-reported incidents Information from the WHO SSA was included for 11 countries and territories: Afghanistan, Burkina Faso, the DRC, Libya, Mali, Myanmar, Nigeria, the oPt, Somalia, South Sudan and Yemen. We accessed the SSA on 7 April 2021 for Afghanistan, 24 March for Nigeria and 18 March for the oPt, and included the information for incidents in these countries reported in 2020 that were available on these dates. For all other countries, the SSA was accessed on 15 January 2021. Any changes to the SSA system after that date are not reflected in the SHCC dataset, but may be noted in the country profiles. We coded 229 SSA-reported incidents from the 11 countries and territories based on the information included on the online SSA dashboard. Since the SSA does not provide information on perpetrators, we assumed that all of the SSA incidents we included involved conflict actors (rather than private individuals) and therefore fulfilled the SHCC inclusion criteria. The SSA also does not provide any information on location, except for the country where the incident occurred. The SSA-reported incidents could therefore not be included in the maps showing the affected regions or provinces in the individual country profiles. The lack of detail in the 28 SSA-reported incidents from Syria made it too difficult to determine which of these incidents overlapped with the 121 Syrian incidents collected by SHCC members. Thus, the 28 SSA-reported incidents from Syria were not incorporated into the report. The SSA includes the fields of ‘Affected Health Resource’, ‘Type of Attack’, and ‘Affected Personnel’, with standard categories for each incident. However, these fields were not consistently filled in, and for 35 of the 229 incidents only one or two of the fields provided information. When one or more fields were left empty, it was usually not possible to fully understand the nature of the incident from the information reported. Therefore, 35 SSA-reported incidents appear in the SHCC dataset as recorded incidents without much further detail, and 194 incidents reported by the SSA are included with more details.13 Limitations of the research This report is based on a dataset of incidents of violence against health care that has been systemically compiled from a range of trusted sources and carefully coded. The figures presented in the report can be cited as the total number of incidents of attacks on health care in 2020 reported or identified by the SHCC. These numbers provide a minimum estimate of the damage to health care from violence and threats of violence that occurred in 2020. However, the severity of the problem is likely much greater, because many incidents probably go unreported and are thus not counted here. Moreover, differences in definitions and biases within individual sources suggest that the contexts that are identified are also not representative of the contexts of all incidents. NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 8
Methodology The SHCC dataset aims to bring together available information from different sources on violence and threats of violence against health care. As a consequence, it suffers from limitations inherent in the information provided by contributors to the SHCC. For some countries, combining available information is challenging when various data collection efforts do not share data in a way that allows information to be cross-checked. Moreover, not all contributors provided access to their original sources and many details were lost in the process, affecting our ability to provide more accurate and consistent classification. This results in two important warnings: The reported numbers of incidents by country should not be compared to those of other countries without considering the factors that affect the flows of information. For example, the information flows from Syria and the oPt are well established. As a result, a relatively high proportion of incidents are generally reported. For a number of countries that emerged as new concerns in 2020, the SHCC made special efforts to improve related data flows, among them Azerbaijan, Burkina Faso, Cameroon, Mozambique, Myanmar and Somalia, but these information flows need further attention. For some other countries, in particular the Central African Republic (CAR), the flow of information remains very challenging. The reported categories of the contexts in which incidents took place should not be read as describing the full range of particular incidents or how frequently they occur. For example, the killings and kidnappings of doctors or bombings of hospitals are more likely to be captured by reporting systems than the harassment of health workers or looting of medical supplies. These incidents are likely to occur more frequently than reports indicate. Reporting and selection bias The SHCC dataset suffers from ‘reporting bias’, which is the technical term for selective reporting. While the process of data cleaning carried out by the SHCC focuses exclusively on selecting incidents based on the inclusion criteria, the pool of information accessible for this process depends on the work done by those who first reported the incidents. Events may be selected or ignored for a range of reasons, including editorial choices, when the source is a media outlet; lack of knowledge, because the affected communities had no connection to the body compiling the information in the first place; or simple errors of omission. These biases mean that the SHCC’s collection of incidents may not be complete or representative, and that only a selection of incidents is included in the first lists that are used to compile the final SHCC dataset. This dataset therefore only covers a fraction of the relevant evidence and covers incidents in certain countries and certain types of incidents more widely than others. NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 9
Methodology
Known reporting and selection biases in SHCC sources
The dataset on which this report is based suffers from the limitations inherent in the
contributors’ data sources used to compile the dataset. Some data sources use media
reports, while others collect and collate reports through a network of partners, direct
observation or the triangulation of sources. Many information providers use a combination
of these methods. Seven possible reporting biases affect the flow of information:
1. In some countries the media frequently report a wide range of attacks on health care,
while in others formal media outlets report hardly any incidents.
2. In some countries citizen journalists who carry out their own documentation and
investigations are key sources of information. Government-imposed shutdowns of the
internet can disrupt such information flows during specific time periods.
3. In some countries there are very active networks of SHCC partner organizations who
contribute information, while in others no such networks exist. Building up networks
takes time and these networks are better developed in countries experiencing long-
standing conflicts. Changes in personnel or funding shortfalls can disrupt information
flows.
4. In some countries numerous parallel data-collection processes exist that publish
different numbers because of differences in geographic coverage or the ability to reach
information providers. Where the original data is not shared, it is impossible to cross-
check for double reporting of the same events.
5. In some countries data collection initiatives may publish data in one year that leads to a
sudden rise in reported incidents. If they do not continue this work in subsequent years,
the numbers of reported incidents then drop.
6. Incidents occurring in the early stages of conflicts need to be found in a variety of sources
until data-collection networks are established.
7. Some organizations do not share incidents in order to protect their independence and
neutrality. In countries where such organizations are key health care providers,
information flows can remain very limited.
Accuracy of information and differing definitions
Some organizations record only certain types of incidents, e.g. those involving health facilities or
those affecting international aid agencies, while the incident descriptions that are available may
also contain errors. In addition, not all organizations that compile information on relevant incidents
include all the details that would be necessary to systematically code all aspects of these incidents.
In particular, information related to the perpetrator(s) and context of a particular incident is often
missing or may be biased in the original source. Also, in some cases, especially those involving
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 10Methodology
robberies and abductions, it is often difficult to ascertain from available information whether the
act was committed by a party to the conflict or by criminals. We based our inclusion decisions on
judgements about the most likely motivations.
The nature of the WHO SSA dataset and the extent to which the SHCC relies on contributions
from this dataset for specific countries influence the overall SHCC dataset. Because the SSA does
not report information on perpetrators, the SHCC dataset could not provide information on the
perpetrators in 229 incidents. As a consequence, the coding is much more limited for those
countries for which a significant proportion of incidents came from the SSA. In addition, the SSA
reported 35 incidents that did not contain enough precise information to include the events in
the SHCC dataset beyond the incident count.
The SHCC dataset therefore contains limitations associated with using preprocessed data without
access to the original sources or additional detail, which would have allowed for potentially more
comprehensive and consistent classification.
The standard coding principles are set out in the SHCC Overview Data Codebook. Please see
www.insecurityinsight.org/projects/healthcare/shcc for full details regarding SHCC coding and
annexes.
2 Department of Peace and Conflict Research, Uppsala University. Uppsala Conflict Data Program. https://ucdp.uu.se/
(accessed 6 April 2021).
3 Department of Peace and Conflict Research, Uppsala University. UCDP Definitions. https://www.pcr.uu.se/research/ucdp/
definitions/.
4 https://ucdp.uu.se/. Because the 2020 UDCP country conflict list was not publicly available when this report was being
written, we consulted UCDP staff via email to obtain information on the changes related to countries included in the
UCDP list for 2020.
5 http://insecurityinsight.org/projects/health care/monthlynewsbrief.
6 https://aidworkersecurity.org/.
7 https://airwars.org/.
8 http://sn4hr.org/.
9 https://civilianimpactmonitoring.org/.
10 https://www.acleddata.com/.
11 https://www.map.org.uk/.
12 https://www.sams-usa.net/.
13 Please contact Insecurity Insight if you would like more details on the process of including SSA-reported incidents in the
SHCC dataset.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 11The Safeguarding Health in Conflict Coalition is a group of more than 40 organizations
working to protect health workers and services threatened by war or civil unrest. We
have raised awareness of global attacks on health and pressed United Nations
agencies for greater global action to protect the security of health care. We monitor
attacks, strengthen universal norms of respect for the right to health, and demand
accountability for perpetrators. www.safeguardinghealth.org.
Safeguarding Health in Conflict Coalition
615 N. Wolfe Street, E7143, Baltimore, MD 21205
Jenny Jun, cphhr@jhu.edu
Suggested citation: Safeguarding Health in Conflict Coalition (SHCC) 2021.
No Respite: Violence against Health Care in Conflict. May 2021.
NO RESPITE: VIOLENCE AGAINST HEALTH CARE IN CONFLICT | 2020 12You can also read
